Optimization-based damage detection in truss and frame structures using vibration data

Structural health monitoring is of great importance to ensure that structures are safe, durable, and long-service life. Therefore, it is necessary to periodically inspect possible damages under the loads and environmental effects to which structures are exposed throughout their service life. In this context, although damage detection methods based on vibration characteristics using optimization algorithms have gained popularity, they are still among the innovative approaches being researched. In this study, an optimization-based damage detection method using vibration characteristics for truss and frame structural systems is proposed. In the approach, the structures are modeled by SAP2000, and the modal analysis results are transferred to the MATLAB using the SAP-OAPI feature. Teaching-Learning Based Optimization (TLBO) algorithm is used to minimize the objective function and match the calculated vibration characteristics of damaged structure with those of the reference (undamaged) structure. As a result of the analysis, the damaged members and the extent of the damages are determined. The method is tested on a 47-element steel planar truss, a 36-element steel space truss, and 24-element three-dimensional reinforced concrete frame system with and without noisy conditions. The obtained results indicate that the proposed method provides sufficient accuracy and reliability in damage detection. It has been evaluated that damage detection methods based on vibration data can be used as an effective tool in ensuring the safety of structures.